Age-related hearing loss (HL), a common and treatable condition, has been associated with other age-related conditions. Late life cognitive impairment is a major public health concern that is rarely treatable. Studies examining the relationship between HL and cognition have been limited by non-Hispanic cohorts, small samples, or limited confounding control. We overcome these limitations in a large Hispanic cohort. This was a multisite cross-sectional study of 5,277 subjects at least 50 years old (Hispanic Community Health Study, HCHS). The main exposure was audiometric HL. The main outcome measure was neurocognitive performance ascertained by the Digit Symbol Substitution Test (DSST), Word Frequency Test, Spanish-English Verbal Learning Test (SEVLT), and Six-Item Screener. The mean age was 58.4 years (SD = 6.2). A 20-dB (equivalent to a one-category worsening) increase in HL was associated with a -1.53 (95% CI, -2.11, -0.94) raw score point difference in the DSST, adjusting for demographics, hearing aid use, and cardiovascular disease. Similarly, a 20-dB increase in HL was associated with a -0.86 (-1.23, -0.49) point difference on the Word Frequency Test, -0.76 (-1.04, -0.47) on the SEVLT 3 trials, -0.45 (-0.60, -0.29) on the SELVT recall, and -0.07 (-0.12, -0.02) on the Six-Item Screener. In the largest study of formal, audiometric HL and cognition to date, HL was independently associated with worse performance in a range of neurocognitive measures. Because HL is common and potentially treatable, it should be investigated as a modifiable risk factor for neurocognitive decline and dementia.